(η6-Arene)osmium(II) Complexes - American Chemical Society

Osmium-187 NMR data have been collected for 37 Os(arene)X2L type ... There are some correlations between 1J(187Os,31P) and the acceptor character...
0 downloads 0 Views 444KB Size
Download PDF
3124

Organometallics 1996, 15, 3124-3135

187Os

NMR Study of (η6-Arene)osmium(II) Complexes: Separation of Electronic and Steric Ligand Effects† Andrew G. Bell, Wiktor Koz´min´ski, Anthony Linden, and Wolfgang von Philipsborn*,‡ Organisch-chemisches Institut, Universita¨ t Zu¨ rich-Irchel, Winterthurerstrasse 190, CH-8057 Zu¨ rich, Switzerland Received January 26, 1996X

Osmium-187 NMR data have been collected for 37 Os(arene)X2L type complexes, using inverse two-dimensional (31P, 187Os){1H} and (1H, 187Os) NMR spectroscopy. In the series Os(p-cymene)X2PMe3, shielding increases in the order Cl < Br < I < Me < H. Systematic variation of the phosphite ligand of Os(p-cymene)Cl2P(OR)3 (R ) Me, Et, nBu, Ph, iPr) reveals a linear dependence on Tolman’s electronic parameter χ with metal shielding increasing as the acceptor character of the P(OR)3 ligand increases. Conversely, the structurally related Os(p-cymene)X2PR3 types (X ) Cl, I; PR3 ) PMe3, PMe2Ph, PnBu3, PPh2Me, P(CH2Ph)3, PPh3, P(m-tolyl)3, PiPr3, PCy3) exhibit a linear dependence on Tolman’s cone angle θ, with shielding decreasing as larger phosphine ligands are introduced. Variation of the arene ligand of Os(C6H5X)Cl2PMe3 (X ) H, Me, Et, iPr, tBu) shows a linear dependence of chemical shift on Taft’s steric parameter Es for group X. In only one case, for the monohydride complexes Os(p-cymene)HClPR3 (PR3 ) PMe3, PMe2Ph, PnBu3, PPh3, PiPr3, PCy3), has it been possible to separate both steric and electronic effects of the phosphorus ligand on the observed 187Os chemical shifts. There are some correlations between 1J(187Os,31P) and the acceptor character of the phosphorus ligand, and a relationship between Os-P bond lengths and coupling constants was found. Osmium relaxation rates T1-1 and their B0 field dependence are reported for selected complexes. The crystal structures of four Os(p-cymene)Cl2L complexes (L ) PMe3 (1), P(OMe)3 (6), PBz3 (13), and PPh3 (15)) have been determined. Introduction Transition-metal NMR spectroscopy can be used to explore both electronic2 and steric3 effects in organometallic systems. In addition, its use as a probe to screen the activity of catalysts,4 correlate rate data,5 and provide insight into mechanistic pathways6 is an ongoing area of research in our group, although these attempts have not always been successful.7 Given the current interest in osmium for catalytic transformations,8 we decided to examine this nucleus in more detail †

Transition Metal NMR Spectroscopy. 32. Part 31: Reference 1. E-mail: [email protected]. Abstract published in Advance ACS Abstracts, June 1, 1996. (1) Meier, E. J. M.; Koz´min´ski, W.; Lustenberger, P.; Linden, A.; von Philipsborn, W. Organometallics 1996, 15, 2469. (2) Graham, P. B.; Rausch, M. D.; Ta¨schler, K.; von Philipsborn, W. Organometallics 1991, 10, 3049. (3) (a) Asaro, F.; Costa, G.; Dreos, R.; Pellizer, G.; von Philipsborn, W. J. Organomet. Chem. 1996, 513, 193. (b) Ludwig, L.; O ¨ hrstro¨m, L; Steinborn, D. Magn. Reson. Chem. 1995, 33, 984. (c) Elsevier, C. J.; Kowall, B.; Kragten, H. Inorg. Chem. 1995, 34, 4836. (d) Tavagnacco, C; Balducci, G; Costa, G.; Ta¨schler, K; von Philipsborn, W. Helv. Chim. Acta 1990, 73, 1469. (4) (a) Bo¨nnemann, H.; Brijoux, W.; Brinkmann, R.; Meurers, W.; Mynott, R.; von Philipsborn, W.; Egolf, T. J. Organomet. Chem. 1984, 272, 231. (b) Bender, B. R.; Koller, M.; Nanz, D.; von Philipsborn, W. J. Am. Chem. Soc. 1993, 115, 5889. (5) (a) Koller, M.; von Philipsborn, W. Organometallics 1992, 11, 467. (b) DeShong, P.; Sidler, D. R.; Rybczynski, J.; Ogilvie, A. A.; von Philipsborn, W. J. Org. Chem. 1989, 54, 5432. (c) DeShong, P.; Slough, G. A.; Sidler, D. R.; Rybczynski, J.; von Philipsborn, W.; Kunz, R. W.; Bursten, B. E.; Clayton, T. W. Organometallics 1989, 8, 1381. (6) Tedesco, V.; von Philipsborn, W. Organometallics 1995, 14, 3600. (7) Dowler, M. E.; Le, T.-X.; DeShong, P.; von Philipsborn, W.; Vo¨hler, M.; Rentsch, D. Tetrahedron 1993, 49, 5673. (8) For a recent review of homogeneous catalysis by osmium complexes, see: Sa´nchez-Delgado, R. A.; Rosales, M.; Esteruelas, M. A.; Oro, A. J. Mol. Catal. A 1995, 96, 231. ‡

X

S0276-7333(96)00053-2 CCC: $12.00

as part of our research on the application of transitionmetal NMR spectroscopy. Osmium has two magnetically receptive nuclei, although the 189Os isotope is not useful because it has a spin of 5/2 and contains a severe quadrupole moment (Q ) 0.91). The 187Os isotope has spin 1/2 but has a natural abundance of only 1.64% and is the most insensitive nucleus in the periodic table, making its observation by conventional NMR techniques extremely difficult.9 For many years, the only known chemical shift was of the reference compound OsO4.10 However, the advent of polarization transfer techniques, which require a nonzero scalar J(M,X) coupling, has improved the detection of low-γ spin 1/2 nuclei, with sensitivity gains over direct metal detection proportional to (γX/ γM). Further enhancements are possible by employing indirect 2D spectroscopy, which involves detection of the metal resonance at the frequency of a more receptive spin 1/2 nucleus. This gives a factor of (γX/γM)5/2 increase in sensitivity over direct detection, which for 187Os NMR spectroscopy corresponds to gains of 1300 and 12 000 for 31P and 1H-detected spectra. These techniques have been applied to other transition-metal nuclei such as 183W, 103Rh, and 57Fe.11 Mann and Maitlis have successfully applied a double INEPT 2D NMR technique to study some µ-hydrido (9) Mason, J., Ed. Multinuclear NMR; Plenum Press: New York, 1987. (10) Kaufmann, J.; Schwenk, A. Phys. Lett. 1967, 24A, 115. (11) (a) Pregosin, P. S., Ed. Transition Metal Nuclear Magnetic Resonance; Elsevier: Amsterdam, 1991. (b) Mann, B. E. In Annual Report on NMR Spectroscopy; Webb, G. A., Ed.; Academic Press: London, 1991; Vol. 23, p 141. (c) von Philipsborn, W. Pure Appl. Chem. 1986, 58, 513.

© 1996 American Chemical Society

(η6-Arene)osmium(II) Complexes Chart 1. Types of Os(arene)X2L Complexes Investigated

binuclear osmium complexes.12 However, the first systematic approach to gaining some δ(187Os), J(Os,X), and T1(187Os) parameters was made more recently by Benn and co-workers for a range of Os(Cp)L2R complexes (Cp ) η5-C5H5), using the HMQC technique.13 The success of this type of spectroscopy has made it the method of choice for determining δ(187Os) and J(Os,X) scalar couplings (X ) H, F, P) in later work.14,15 Results and Discussion An 187Os NMR study has been made of the six classes of structurally related η6-arene-substituted osmium(II) complexes shown in Chart 1.16 Classes I and VI are PMe3 substituted but contain various η1- and η6- ligands, respectively. Classes II and III are dichlorides bearing phosphite and phosphine ligands, class IV compounds are diiodides, and class V compounds are monohydrides. All these types have the advantage of being both soluble, ca. 0.5 M, and stable in CD2Cl2 and THF-d8 solvents, allowing for easier detection of the osmium resonance. X-ray Crystal Structures. Some selected bond angles and lengths for complexes 1, 6, 13, and 15 are given in Table 1, and the crystal structures of 1 (molecule A) and 13 are shown in Figure 1. The structure of 1 has two molecules in the asymmetric unit (A and B); however, there are no major differences in their conformation other than a small rotation of the PMe3 group. The complexes exhibit a number of structural similarities, as well as some noticeable differences. The PR3 ligands of 1, 6, and 13 adopt a (12) (a) Cabeza, J. A.; Mann, B. E.; Maitlis, P. M.; Brevard, C. J. Chem. Soc., Dalton Trans. 1988, 629. (b) Cabeza, J. A.; Nutton, A.; Mann, B. E.; Brevard, C.; Maitlis, P. M. Inorg. Chim. Acta 1986, 115, L47. (c) Cabeza, J. A.; Mann, B. E.; Brevard, C.; Maitlis, P. M. J. Chem. Soc., Chem. Commun. 1985, 65. (13) (a) Benn, R.; Brenneke, H.; Joussen, E.; Lehmkuhl, H.; Ortiz, F. L. Organometallics 1990, 9, 756. (b) Benn, R.; Joussen, E.; Lehmkuhl, H.; Ortiz, F. L.; Rufinska, A. J. Am. Chem. Soc. 1989, 111, 8754. (14) Christe, K. O.; Dixon, D. A.; Mack, H. G.; Oberhammer, H.; Pagelot, A.; Sanders, J. C. P.; Schrobilgen, G. J. J. Am. Chem. Soc. 1993, 115, 11279. (15) Michelman, R. I.; Ball, G. E.; Bergman, R. G.; Andersen, R. A. Organometallics 1994, 13, 869. (16) For a review of the organometallic chemistry of (η6-arene)Os complexes, see: LeBozec, H.; Touchard, D.; Dixneuf, P. H. Adv. Organomet. Chem. 1989, 29, 163.

Organometallics, Vol. 15, No. 14, 1996 3125

staggered conformation, with the Os-P bond bisecting the C(2)-C(3) bond of the arene at approximately its midpoint. It is noticeable with 13 that the iPr group bound to the ring points away from the phosphine ligand, as evidenced by the increase in the C(2)-C(1)C(7)-C(9) torsion angle. However, 15 adopts a conformation in which the Os-P bond bisects the C(4)-C(5) bond distant to the bulky iPr group, presumably for steric reasons, resulting in a restoration of the C(2)C(1)-C(7)-C(9) torsion angle. The p-cymene ligand is tilted toward the OsCl2L moiety, placing the C-C bond that bisects the Os-P bond closer to the metal than the parallel C-C bond [C(5)-C(6) or C(1)-C(2)], but this becomes less marked as larger ligands are introduced. The p-cymene ring is puckered in a manner dictated by the size of the phosphorus ligand. Thus, the χ2 distributions17 for deviations of the η6-arene atoms C(1) to C(6) from planarity are 30.3 for 1b, increasing to 289.5 for 15. For complexes 1 and 13, atoms C(1) and C(4) lie above the average plane of the η6-arene ring, with the others below. However, for the P(OMe)3substituted case 6, the reverse is true. Complex 15 behaves like 6, except that, due to the change in conformation of the PPh3 ligand, it is now the C(3)C(6) axis which lies beneath the plane of the arene ligand. Finally, the crystal structures suggest that as the phosphorus ligand becomes more sterically demanding, the Os-P bond distance increases in length. However, the effect of the ligand on the Os-Cl bond lengths and Cl-Os-Cl bond angles varies, especially for 1. Chemical Shifts. Ramsey18 described nuclear shielding (σ) as arising from a diamagnetic (σd) and paramagnetic (σp) contribution, of which the former remains fairly constant for a given nucleus. Later work by Griffith and Orgel19 has elaborated upon the Ramsey model to give a better description for transition metal nuclei, for which the σp contribution to the nuclear shielding term is shown in eq 1, where ∆E is the average

∑QN

σP ) -constant (∆E)-1〈rd-3〉

(1)

d-d excitation energy and rd is the effective radius of the d-orbital electron. The ∑QN term describes the distribution of charge in the valence p and d orbitals, which can be assumed to remain fairly constant for structurally related molecules. Initially, the concentration and temperature dependence of δ(187Os) was investigated. At 300 K, the shifts of 6 in CD2Cl2 exhibit a linear concentration dependence (six points, concentration coefficient 18.3 ppm‚L‚mol-1), with the osmium resonance at 7.7 × 10-2 M (20 mg in 0.5 mL of solvent) being more highly shielded by 7 ppm than in a 4.6 × 10-1 M solution (120 mg in 0.5 mL). Similarly, a linear temperature coefficient of +0.7 ppm K-1 was observed for a 4.6 × 10-1 M solution of 6 between 233 and 303 K (six points) in polar CD3CN and nonpolar CD2Cl2 solvents. This effect is in the same order as that observed for Fe(CO)520 but is substantially less than the value of 2.48 ppm K-1 (17) Stout, G. H.; Jensen, L. H. X-Ray Structure Determination-A Practical Guide, 2nd ed.; John Wiley & Sons: New York, 1989; p 409. (18) Ramsey, N. F. Phys. Rev. 1950, 77, 567; 78, 699; 1951, 83, 540; 1952, 86, 243. (19) Griffith, J. S.; Orgel, L. E. Trans. Faraday Soc. 1957, 53, 601. (20) Schwenk, A. Phys. Lett. 1970, 31A, 513.

3126

Organometallics, Vol. 15, No. 14, 1996

Bell et al.

Table 1. Selected Bond Lengths (Å), Angles (deg), and Torsion Angles (deg) for Complexes 1, 6, 13, and 15 1

a

molecule A

molecule B

6

13

15

Os-Cl(1) Os-Cl(2)

2.424(2) 2.426(6)

2.428(2) 2.416(1)

2.422(2) 2.425(2)

2.4251(7) 2.4220(7)

2.427(1) 2.414(1)

Os-P

2.341(2)

2.340(2)

2.281(1)

2.3608(7)

2.355(1)

Os-C(1) Os-C(2) Os-C(3) Os-C(4) Os-C(5) Os-C(6) χ2 a

2.201(6) 2.181(6) 2.158(6) 2.184(6) 2.235(5) 2.248(5) 100.5

2.215(6) 2.194(5) 2.186(5) 2.201(5) 2.189(5) 2.236(5) 30.3

2.204(6) 2.180(5) 2.199(6) 2.229(5) 2.270(5) 2.276(5) 145.4

2.212(3) 2.171(3) 2.210(3) 2.225(3) 2.248(3) 2.237(3) 154.7

2.258(4) 2.231(4) 2.187(4) 2.236(4) 2.203(4) 2.169(4) 289.5

Cl(1)-Os-Cl(2)

85.73(6)

88.00(6)

86.27(6)

87.78(3)

87.18(4)

Cl(1)-Os-P Cl(2)-Os-P

83.22(6) 87.94(6)

82.71(5) 85.33(6)

86.37(5) 90.48(6)

85.06(3) 79.69(2)

85.54(3) 87.55(4)

C(2)-C(1)-C(7)-C(8) C(2)-C(1)-C(7)-C(9)

105.1(9) -17.5(9)

110.8(7) -14.5(8)

93.7(7) -31.6(9)

-168.6(3) 68.0(3)

95.1(4) -29.3(5)

Deviation of C(1)-C(2)-C(3)-C(4)-C(5)-C(6) from planarity.17

Figure 1. ORTEP50 plot of Os(p-cymene)Cl2PMe3 (1, molecule A) and Os(p-cymene)Cl2P(CH2Ph)3 (13) (50% probability ellipsoids; H atoms given arbitrary thermal parameters for clarity).

observed for the complex Fe(Cp)(PPh3)(CO)(COMe).21 Therefore, with careful temperature and concentration

control of our samples, we predict that our chemical shifts are accurate to within (1 ppm. The 187Os shifts of compounds 1-37 span a range of approximately 3500 ppm in total but show considerable variations according to class (Tables 2-4). The greatest variation (approximately 3000 ppm) occurs in class I as the electronegativity of X is varied. Changing alkyl substituents on the η6-ligand of VI still has a considerable influence on chemical shift (approximately 600 ppm). The phosphine-substituted examples III and IV exhibit a similar chemical shift range (approximately 550 ppm) while the phosphites II and monohydrides V contain less variation (approximately 150 ppm). By systematically varying the stereoelectronic properties of the ligands in the coordination sphere of the metal, we have been able to establish a number of structural correlations as a function of the osmium chemical shift and have been able to separate electronic and steric effects. I. Electronic Effects. Complexes I exhibit a “normal” halogen dependence, with deshielding in the order Cl > Br > I > sp3 C > H. Thus, replacement of both chlorine ligands of 1 with bromine results in a 270 ppm shielding of the osmium nucleus (moves to lower frequency). Further displacement of bromine by iodine causes an additional 760 ppm shielding, while the dihydride 5 resonates a further 2000 ppm to lower frequency of the diiodide 4 and is the most shielded example in this study. This indicates that the observed shielding effects are largely determined by the 〈rd-3〉 term in eq 1.22 These ligand effects largely parallel those reported by Benn et al.13a for Os(C5H5)(PR3)2X complexes. The shifts of complexes II (Figure 2) exhibit a good linear relationship (r ) 0.96) with Tolman’s electronic parameter χ for the P(OR)3 ligand,23 with the metal nucleus becoming progressively more shielded with increasing π-acidity (larger χ values). However, with the exception of the P(OMe)3 complex 6, type II phosphite complexes are more shielded relative to their phosphine counterparts of series III. This “inverse” (21) Meier, E. J. M.; von Philipsborn, W. Unpublished results. (22) Mason, J. Chem. Rev. 1987, 87, 1299. (23) Bartik, T.; Himmler, T.; Schulte, H. G.; Seevogel, K. J. Organomet. Chem. 1984, 272, 29.

(η6-Arene)osmium(II) Complexes Table 2. no.

187Os

Organometallics, Vol. 15, No. 14, 1996 3127 and

31P

NMR Data (CD2Cl2, 300 K) for Type I and II Complexes δ(187Os)/ppm

compd

δ(31P)/ppm

1J(Os,P)/Hz

nJ(Os,H)/Hz

-39.2 -48.3 -61.6 -41.8 -44.5a

272 269 266 280 264

5 (2J) 79 (1J)

73.6 69.2 68.0 58.4 63.6

453 448 447 477 442

Type I 1 2 3 4 5

Os(p-cymene)Cl2PMe3a Os(p-cymene)Br2PMe3 Os(p-cymene)I2PMe3 Os(p-cymene)ClMePMe3 Os(p-cymene)H2PMe3b

-2226 -2495 -3260 -3330 -5265a

Os(p-cymene)Cl2P(OMe)3 Os(p-cymene)Cl2P(OEt)3 Os(p-cymene)Cl2P(OnBu)3 Os(p-cymene)Cl2P(OPh)3 Os(p-cymene)Cl2P(OiPr)3

-2168 -2142 -2153 -2257 -2076

Type II 6 7 8 9 10 a

p-cymene ) p-MeC6H4CHMe2.

Table 3.

a

b

Measured in THF-d8.

187Os

and

31P

NMR Data (CD2Cl2, 300 K) of Type III-V Complexes δ(31P)/ppm

1J(Os,P)/Hz

Type III -2226 -2162 -2097 -2028 -2081 -1906 -1892 -1735 -1697

-39.2 -33.9 -20.3 -22.0 -26.6 -13.1 -14.3 -8.5 -18.7

272 271 275 278 270 282 281 269 266

Os(p-cymene)I2PMe2Ph Os(p-cymene)I2PnBu3 Os(p-cymene)I2PPh2Me Os(p-cymene)I2PPh3 Os(p-cymene)I2PiPr3 Os(p-cymene)I2PCy3

Type IV -3201 -3119 -3173 -3036 -2738 -2688

-54.6 -46.1 -37.6 -23.7 -21.7 -31.2

268 268 275 285 270 268

Os(p-cymene)ClHPMe3 Os(p-cymene)ClHPMe2Ph Os(p-cymene)ClHPnBu3 Os(p-cymene)ClHPPh2Me Os(p-cymene)ClHPPh3 Os(p-cymene)ClHPiPr3 Os(p-cymene)ClHPCy3

Type Va -3758 -3723 -3814 -3740 -3680 -3764 -3738

-40.3 -27.8 -9.1 -7.8 -8.2 22.5 14.5

260 264 263 270 277 265 264

no.

compd

1 11 12 13 14 15 16 17 18

Os(p-cymene)Cl2PMe3 Os(p-cymene)Cl2PMe2Ph Os(p-cymeneCl2PPh2Me Os(p-cymene)Cl2P(CH2Ph)3 Os(p-cymene)Cl2PnBu3 Os(p-cymene)Cl2PPh3 Os(p-cymene)Cl2P(m-tolyl)3 Os(p-cymene)Cl2PiPr3 Os(p-cymene)Cl2PCy3

19 20 21 22 23 24 25 26 27 28 29 30 31

δ(187Os)/ppm

1J(Os,H)/Hz

73 72.5 72 71 70 69 68.5

Measured in THF-d8.

Table 4.

187Os

and 31P NMR Data (CD2Cl2, 300 K) for Type VI Complexes

no.

compd

δ(187Os)/ ppm

δ(31P)/ ppm

32 33 34 35 36 1 37

Os(C6H6)Cl2PMe3 Os(C6H5-Me)Cl2PMe3 Os(C6H5-Et)Cl2PMe3 Os(C6H5-iPr)Cl2PMe3 Os(C6H5-tBu)Cl2PMe3 Os(p-cymene)Cl2PMe3 Os(C6Me6)Cl2PMe3

-2431 -2364 -2352 -2348 -2268 -2226 -1829

-37.5 -37.7 -38.0 -38.2 -38.0 -39.2 -42.3

1J(Os,P)/

Hz 266 269 270 270 267 272 290

ligand behavior is probably due to an increase in the spectrochemical ∆E parameter caused by a larger metal-ligand orbital interaction of more strongly πaccepting P(OR)3 ligands. This offsets the deshielding effect expected by way of the associated decrease in the nephelauxetic rd parameter. An analogous effect was reported for the osmium bis(phosphine) and bis(phosphite) complexes.13a II. Steric Effects. The shifts of the phosphinesubstituted compounds III are linearly dependent on Tolman’s steric parameter θ for the PR3 ligand24 (r ) 0.98) with the osmium nucleus becoming progressively more deshielded as larger phosphine ligands are intro(24) Tolman, C. A. Chem. Rev. 1977, 77, 313.

duced (Figure 3). Similarly, the iodides IV, which are on average 1050 ppm more shielded than III, give a linear correlation between δ(187Os) and θ (r ) 0.97). There appears to be a mild ligand effect, as the correlation is less significant than for the analogous dichloride complexes III. This effect has been observed for some (Cp*)RhX2PR3 complexes (X ) Cl, Br, I, Cp* ) η5C5Me5).25 However, a plot of the osmium shifts of III against IV is linear (r ) 0.98), suggesting that the introduction of progressively larger PR3 ligands in both cases has the same and co-operative effect of decreasing both the ∆E and rd parameters because of their size.26 Similar observations have been made by Elsevier for some (η4-COD)Rh(PR3)Cl complexes.3c The osmium shifts of complexes VI reveal a deshielding influence as more sterically demanding arene groups are introduced. Thus, the hexamethylbenzene-substituted complex 37 resonates 600 ppm to higher frequency relative to the parent complex 32. Unfortunately, relatively little is known about the steric or electronic properties of arene ligands. In steric terms, Maitlis has calculated cone angles of 162 and 192° for the Ru-C6H6 (25) Tedesco, V., von Philipsborn W. Magn. Reson. Chem. 1996, 34, 373. (26) Rehder, D. Chimia 1986, 40, 186.

3128 Organometallics, Vol. 15, No. 14, 1996

Figure 2. Plot of δ(187Os) vs Tolman’s electronic parameter χ for complexes II (r ) 0.96).

Figure 3. Plot of δ(187Os) vs Tolman’s steric parameter θ for complexes III (r ) 0.98): (a) cone angles taken from ref 24; (b) cone angle taken from ref 34; (c) cone angle determined from crystal structure of 13.

and Ru-C6Me6 moieties.27 Their electronic properties have been more thoroughly investigated by determining core binding energies in the (Cp)Fe(η6-arene)+ system, using X-ray photoelectron spectroscopy.28 This study has shown that the introduction of a methyl substituent onto the arene slightly enhances its donor properties, although the effect is not great. We have studied chemical shift correlations with Hammett and Taft-type substituent constants for the five Os(η6-C6H5X)Cl2PMe3 complexes 32-36 (X ) H, Me, Et, iPr, tBu). Statistically insignificant correlations were obtained for Hammett σm, σp, σp+, and σp- and Swain-Lupton F and R values,29 whereas an excellent linear correlation (Figure 4) was obtained using Taft’s (27) Maitlis, P. M. Chem. Soc. Rev. 1981, 10, 1. (28) Gassman, P. G.; Deck, P. A. Organometallics 1994, 13, 2890. (29) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.

Bell et al.

Figure 4. Plot of δ(187Os) vs Taft’s steric parameter Es for complexes VI (r ) 0.99).

Figure 5. Plot of δ(187Os) vs Tolman’s electronic parameter χ for complexes V.

steric parameter Es (r ) 0.99).30 This clearly demonstrates the steric origin of the deshielding effects and complements previous observations for 59Co,3d 103Rh,3a,b and 57Fe.1 The poor correlations obtained using the field and resonance constants, F and R, further suggest that the steric influence of the arene substituent X operates by a combination of field and resonance effects. III. Mixed Effects. In one case, it has been possible to observe both the electronic and steric effects of similar ligands within the same system. The monohydrides V are easily accessible from III but give no linear correlation between δ(187Os) and other structural parameters. A plot of δ(187Os) against χ (Figure 5) does reveal a nearlinear correlation with the ligands PnBu3 to PPh3, but unlike II the metal shifts exhibit a “normal” ligand dependence and are shielded by less π-acidic ligands. This behavior changes when larger and less electroni(30) March, J. Advanced Organic Chemistry, 4th ed.; John Wiley & Sons: New York, 1992; p 285.

(η6-Arene)osmium(II) Complexes

Organometallics, Vol. 15, No. 14, 1996 3129

cally demanding phosphines (PiPr3 and PCy3) are introduced. These ligands possess virtually no πacceptor ability and presumably deshield the metal nucleus because of their size, as with III and IV. Thus, it appears that as the metal becomes more electron rich, the acceptor character of the phosphorus ligand becomes the principal factor determining the chemical shift of the osmium nucleus, with better donor ligands shielding the metal, until a threshold of χ is attained (χ < 5.25, θ > 145°) where steric factors become dominant. For comparison, we have attempted similar correlations using more conventional 31P NMR methods. For complexes II and III, χ or θ do not correlate with their absolute 31P shifts but do show modest (r ) 0.93) and good (r ) 0.98) correlations, respectively, with their coordination chemical shifts ∆ (δfree ligand - δcomplex). In contrast, for types IV-VI χ or θ do not correlate with either δ or ∆(31P), thus reflecting a greater sensitivity of the metal nucleus to subtle stereoelectronic changes. We have been able to use the correlation shown in Figure 3 to determine qualitatively the cone angles of various phosphine ligands by 187Os NMR. For example, literature values for P(CH2Ph)324,31 and P(m-tolyl)332 suggest equivalent cone angles of 165°, which do not correspond with the observed shifts of complexes 13 and 16. However, these values were measured using simple CPK molecular models with rotation of the R groups to achieve C3v symmetry.33 Coville has calculated a cone angle of 148° for P(m-tolyl)3 on the basis of a nonsymmetrical arrangement of the three phenyl rings.34 Evidence for this geometry is now provided by a variable-temperature 1H NMR investigation of the m-tolyl region of 16 (Figure 6). At room temperature, all three methyl substituents are equivalent, giving a single resonance. However, at -50 °C the signal broadens and is resolved at -70 °C into two separate resonances with a 2:1 ratio, indicating nonequivalence of the m-tolyl rings. Similarly, the crystal structure of complex 13 (Figure 1) reveals a different orientation of the three benzyl groups. One phenyl ring is directed toward the p-cymene group with the other two pointing away, probably to reduce steric interactions about the congested metal center. This arrangement gives a cone angle 5-6° smaller than that measured for the PPh3 complex 15.35 These corrected cone angles agree with the observed 187Os shifts obtained for complexes 13 and 16. Coupling Constants. A simplified expression to describe one-bond metal-phosphorus coupling constants is given in eq 2,36 where |ΨM(0)|2 and |ΨP(0)|2

JM,P ∝ KRM2RP2|ΨM(0)|2|ΨP(0)|2/3∆E

(2)

are s-electron densities at the nucleus of each atom, R2 represents the localized hybrid bond s-character in a valence bond description, and K is the reduced coupling constant. (31) Manzer, L. E.; Tolman, C. A. J. Am. Chem. Soc. 1975, 97, 1955. (32) (a) Liu, H. Y.; Eriks, K.; Prock, A.; Giering, W. P. Organometallics 1990, 9, 1758. (b) Rahman, M. M.; Liu, H. Y.; Prock, A.; Giering, W. P. Organometallics 1987, 6, 650. (33) Tolman, C. A. J. Am. Chem. Soc. 1970, 92, 2956. (34) (a) White, D.; Coville, N. J. Adv. Organomet. Chem. 1994, 36, 95. (b) White, D.; Carlton, L.; Coville, N. J. J. Organomet. Chem. 1992, 440, 15. (35) Cone angles were measured using the equation given in ref 24 with the Os-P bond set to 2.28 Å. (36) Pregosin, P. S.; Kunz, R. W. 31P and 13C NMR of Transition Metal Phosphine Complexes; Springer-Verlag: Berlin, 1979.

Figure 6. Variable-temperature 1H-NMR spectra (1.92.7 ppm region) of Os(p-cymene)Cl2P(m-tolyl)3 (16) in CD2Cl2 at 400 MHz showing two methyl resonances for the P(m-tolyl)3 ligand at 203 K.

Values of 1J(M,P) scalar coupling constants are known to depend on a number of contributing factors, such as the electronegativity of substituents on phosphorus, metal oxidation state, and structural effects.9,36 However, on the basis of the assumption that 1J(187Os,31P) coupling constants are a reflection of the σ-bond order of the M-P bond, we have attempted to correlate these with the spectral properties of the ligands in the metal coordination sphere. The magnitude of 1J(187Os,31P) in series I increases in the order H < I < Br < Cl < Me, but we do not believe that this trend necessarily reflects the Os-P bond strength. It is generally accepted that M-C bonds are weaker than M-H bonds by 15-25 kcal mol-1, which may genuinely account for the large 187Os,31P coupling obtained for 4. However, thermochemical studies of Ru(Cp*)(PMe3)2X (X ) Cl, Br, I, H) show that the Ru-PMe3 bond dissociation energy increases in the order Cl < Br ≈ I < H.37 A more convincing case is provided by compounds 32-37, which, with the exception of the tBu-substituted complex, show a clear trend of increasing 1J(187Os,31P) with the introduction of larger η6-arene ligands. This observation complements the trend observed with δ(187Os) and clearly suggests a weakening of the osmium-arene bond by larger arene groups, which is also evident from the concomitant strengthening of the Os-P bond. For types II and III, 187Os,31P coupling constants are linearly dependent on the electronic character of the phosphorus ligand χ, giving a very modest correlation (r ) 0.86) for the phosphine series III, which improves (37) Bryndza, H. E.; Domaille, P. J.; Paciello, R. A.; Bercaw, J. E. Organometallics 1989, 8, 379.

3130

Organometallics, Vol. 15, No. 14, 1996 Table 5.

187Os

Bell et al.

Spin-Lattice Relaxation Rates T1-1 (CD2Cl2, 300 K) of Selected Complexesa T1-1/s-1

a

T1-1(csa) (%)

no.

compd

9.4 T

14.1 T

9.4 T

14.1 T

T1-1 (other) 9.1 T

1 2 3 6 7 8 9 10 12 14 15 17 18

Os(p-cymene)Cl2PMe3 Os(p-cymene)Br2PMe3 Os(p-cymene)I2PMe3 Os(p-cymene)Cl2P(OMe)3 Os(p-cymene)Cl2P(OEt)3 Os(p-cymene)Cl2P(OnBu)3 Os(p-cymene)Cl2P(OPh)3 Os(p-cymene)Cl2P(OiPr)3 Os(p-cymene)Cl2PPh2Me Os(p-cymene)Cl2PnBu3 Os(p-cymene)Cl2PPh3 Os(p-cymene)Cl2PiPr3 Os(p-cymene)Cl2PCy3

0.60 0.46 0.39 0.51 0.67 0.89 0.94 0.71 0.88 1.3 1.2 0.87 1.5

1.3 0.98 0.78 1.1 1.6 2.0 2.2 1.7 2.0 3.1 2.7 2.0 3.3

0.56 (93) 0.42 (91) 0.31 (79) 0.47 (92) 0.74 (100)* 0.89 (100) 1.0 (100)* 0.79 (100)* 0.9 (100)* 1.4 (100)* 1.2 (100) 0.9 (100)* 1.4 (100)

1.4 (100)* 0.94 (96) 0.7 (90) 1.1 (100) 1.7 (100)* 2.0 (100) 2.3 (100)* 1.8 (100)* 2.0 (100)* 3.2 (100)* 2.7 (100) 2.0 (100) 3.2 (100)

0.04 0.04 0.08 0.04 0.0* 0.03 0.0* 0.0* 0.0* 0.0* 0.0 0.0* 0.1

Asterisk indicates result of calculation of the CSA relaxation rate greater than 100%, but within limit of experimental error.

greatly (r ) 0.99) for the phosphites II. The magnitude of 1J(187Os,31P) increases with increasing π-acidity of the ligand, which also reflects an increase in the electronegativity of group R. The trialkyl-containing examples of III (R ) Me, nBu, iPr, Cy, CH2Ph) exhibit a better correlation against χ with 1J(187Os,31P) generally decreasing with increasing ligand size. However, this behavior deviates considerably when aryl groups are included, as with PPh2Me, possibly indicating variations in π-bonding interactions for these complexes that are not adequately represented by literature values of χ. Opening of the R-P-R bond angle should increase the phosphorus s character in the P-R bond and decrease it in the Os-P bond. Thus, larger phosphorus ligands should possess smaller 187Os,31P couplings, but this is not generally observed within a particular series. These results complement the observation made by Nolan for the cleavage of the dimer [Ru(p-cymene)Cl2]2 using various phosphorus ligands. Here, the enthalpy of the reaction was found to increase in the order PCy3 < P(OPh)3 < PiPr3 < PPh3 < P(CH2Ph)3 < PPh2Me < P(OMe)3 < PMe2Ph < PMe3 but with a combination of both electronic and steric factors playing an important role in dictating the magnitude of -∆H for the reaction.38 Assuming that the R2 term in eq 2 makes the dominant contribution to 1J(187Os,31P), the correlations observed against χ for II and III suggests that larger coupling constants may be a consequence of shorter M-P bond lengths, as has been shown for some platinum complexes.39 We have compared these couplings with our crystallographic data for the phosphines 1, 13, and 15 and phosphite 6. Although the data are not extensive, it does suggest a relationship in our case (r ) 0.96), although the Os-PMe3 bond of 1 is slightly shorter than the 187Os,31P coupling would predict. The diiodides IV exhibit no similar correlations against χ or θ. This may be an effect of the soft iodine ligand, which can be polarized more easily by the metal to compensate for variations in the bonding requirements of the phosphine. The monohydrides V exhibit no linear correlations between 1J(187Os,31P) and other structural parameters. However, values of 1J(187Os,1H) were found to give a good correlation with Tolman’s steric parameter θ (r ) 0.99), with coupling constants decreasing with increasing phosphine size. Because infrared measurements of these and related complexes40 indicate that the Os-H bond strength increases as larger PR3 ligands are

introduced, we believe that the correlation is not related to the strength of the Os-H bond. It possibly reflects changes of the H-Os-Cl bond angle to accommodate the ligands across the series PMe3 (θ ) 118°) to PCy3 (θ ) 170°). In conclusion, at least two independent effects (bond distances and angles) can influence the value of one-bond 187Os,31P couplings. Therefore, care has to be taken in any interpretation of this parameter in terms of bond strength. Relaxation Times. We have been able to measure accurate 187Os T1 relaxation times for representative examples of types I-III (Table 5).41 Measurements taken of II and III at 9.4 and 14.1 T show that the rate of relaxation varies from 0.39 to 3.3 s-1 and is proportional to the square of the ratio of the two magnetic field strengths (2.25). Hence, the mechanism is chemical shift anisotropy (CSA) dominated. We have compared these T1 rates with other NMR and structural parameters, but only rough correlations with the molecular weight of the attached phosphorus ligand are apparent, with heavier ligands relaxing the metal nucleus more efficiently than smaller ones within their respective groups. This reflects the molecular tumbling or correlation time, τc, of the molecule. However, the phosphites II relax at a rate 15-30% slower than their lighter counterparts of III, which suggests that other factors are involved. This is also evident in the halogen series Os(p-cymene)X2PMe3 (X ) Cl, Br, I), which shows a trend of slower T1 rates as the weight of the halogen increases from chlorine to iodine. This is opposed to an effect of τc and may be related to the polarization of the Os-X bond, which for more electronegative substituents is greater, thus increasing the anisotropy of the ligand field and of the shielding tensor. Further evidence is provided by the fact that the CSA contribution to T1 is substantially reduced at the lower field strength. We are currently trying to extend these measurements to more varied types of complexes to determine whether we can extract additional structural information from this NMR parameter. (38) Serron, S. A.; Nolan, S. P. Organometallics 1995, 14, 4611. (39) Mather, G. G; Pidcock, A.; Rapsey, G. J. N. J. Chem. Soc., Dalton Trans. 1973, 2095. (40) Bennett, M. A.; Weerasuria, M. M. J. Organomet. Chem. 1990, 394, 481. (41) Koz´min´ski, W.; von Philipsborn, W. J. Magn. Reson. A 1995, 116, 262.

(η6-Arene)osmium(II) Complexes

Conclusions Osmium-187 chemical shifts appear to be very sensitive to subtle changes in the coordination sphere of the metal, making this technique useful in gauging both the electronic and steric effects of various ligands. Some of the electronic ligand effects are similar to those observed in bis(phosphine) complexes of the type OsCp(PR3)2X which, however, are characterized by much higher osmium shielding. In addition, osmium-phosphorus and osmium-hydride coupling constants can provide useful structural information in the Os(arene)X2L series, in particular on Os-P bond lengths. It is now possible to run 187Os NMR spectra on a routine basis, and we are currently applying the technique to screen the activity of some homogeneous catalysts. Experimental Section Inverse-detected 187Os NMR spectra were obtained in CD2Cl2 at 300 K using the HMQC technique, with 31P detection at 9.4 T for types I-IV and (VI) and 1H detection at 14.1 T for V. Approximately 120 mg of sample in 0.5 mL of solvent in a 5-mm NMR tube was required to obtain a phosphorus-detected spectrum using a Bruker AM-400-WB spectrometer equipped with a TBI 1H, 31P, BB (9-41 MHz) probe head (ν0(187Os) ) 9.2 MHz). A second PTS160 synthesizer together with a BSV3 decoupling unit was used for pulsing the osmium frequency. Typically 128 increments, each of 24 scans, were accumulated over 3 h to obtain an overview spectrum: spectral width in F1 30 000 Hz, spectral width in F2 800-1000 Hz, 90° pulse length for osmium 60 µs, 90° phosphorus pulse 14.5 µs, relaxation delay 2 s. Final spectra of 256 increments, each of 40 scans, with a spectral window of 2000 Hz in F1 were measured overnight to ensure that the osmium signals were not folded. Proton-detected spectra were acquired on a Bruker AMX-600 spectrometer, which required only 40 mg of sample, using a TBI-LR 1H, 31P, BB (7-21.5 MHz) probe head (ν0(187Os) ) 13.8 MHz). Typically 128 increments, each of 8 or 16 scans, were accumulated in 1 to 2 h: spectral width in F1 800 Hz, spectral width in F2 300 Hz, 90° pulse length for osmium 7080 µs, 90° for proton 11 µs, relaxation delay 2 s. Pulse sequences reported by Benn et al.42 and by Bax et al.43 were used. Osmium-187 spectra of IV were obtained using approximately 20 mg of sample by applying a binomial suppression of the central phosphorus resonance.44 Prior to Fourier transformation, the data were zero filled and multiplied by sine functions in both dimensions. Chemical shifts are reported relative to OsO4 in a field where the protons of TMS resonate at precisely 100 MHz (Ξ ) 2.282 343 MHz).9,10 Osmium relaxation times T1 were measured at the frequency of the 31P nucleus by employing inverse detection methods recently developed in our group.41 Routine 1H and 13C NMR spectra were recorded at 300 K on a Bruker ARX-300 spectrometer and are referenced relative to TMS. 31P NMR spectra were recorded on a Bruker AM400 spectrometer and are referenced relative to 85% H3PO4. For 13C {1H} shift assignments, the p-cymene carbon atoms have been labeled:

All reactions were performed in dry glassware under an inert atmosphere of nitrogen. Solvents were chemically dried (42) Benn, R.; Brevard, C. J. Am. Chem. Soc. 1986, 108, 5622. (43) Bax, A; Griffey, R. H.; Hawkins, B. L. J. Magn. Reson. 1983, 55, 301. (44) Meier, E. J. M.; Koz´min´ski, W.; von Philipsborn, W. Magn. Reson. Chem. 1996, 34, 89.

Organometallics, Vol. 15, No. 14, 1996 3131 and distilled prior to use. The monomeric complexes 1, 3-5, 9, 15, and 29 and dimers [Os(p-cymene)Cl2]2, [Os(C6Me6)Cl2]2, and [Os(C6H6)Cl2]2 were prepared by literature procedures.45-48 Complexes 13, 16, and 18 and osmium trichloride were generously donated by Dr. A. Hafner (Ciba-Geigy, Marly, Switzerland). Os(p-cymene)Br2PMe3 (2). The dichloride 1 (140 mg, 0.3 mmol) and NaBr (300 mg, 3.0 mmol) were stirred in MeOH (10 mL) overnight at room temperature. The solution was evaporated, redissolved in CH2Cl2 (10 mL), and filtered (Celite), and the volume of solvent was reduced to 2-3 mL. Pentane (15 mL) was added to precipitate an orange powder, which was filtered out and washed with cold ether (5 mL). Yield: 150 mg (90%). Mp: 197-209 °C. 1H NMR (CDCl3): p-cymene, δ 1.25 (d, J ) 6.9 Hz, 6H, CHMe2), 2.26 (s, 3H, Me), 2.87 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.45 (m, 4H, MeC6H4CHMe2); PR3, δ 1.72 (d, J ) 10.4 Hz, 9H, P-Me). 13C{1H} NMR (CDCl3): p-cymene, δ 22.3 (s, C7), 18.9 (s, C1), 30.8 (s, C6), 99.1 and 87.9 (s, C2 and C5), 86.1 and 80.6 (s, C3 and C4); PR3, δ 16.8 (d, J ) 39.1 Hz, P-Me). Anal. Calcd for C13H23Br2OsP: C, 27.9; H, 4.11. Found: C, 28.3; H, 4.00. Complexes 6-18 were similarly prepared by cleavage of [Os(p-cymene)Cl2]2 using 2.2 equiv of PR3, of which a representative example is given for the P(OMe)3-substituted complex 6. Os(p-cymeneCl2P(OMe)3 (6). [Os(p-cymene)Cl2]2 (150 mg, 0.2 mmol) and P(OMe)3 (0.5 mL, 0.45 mmol) were stirred in CH2Cl2 (10 mL) for 1 h at 40 °C. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral grade (IV) alumina and diethyl ether as eluant. The solvent was evaporated, and the product was redissolved in CH2Cl2 (2 mL). Pentane (15 mL) was added to precipitate a yellow powder, which was filtered out and washed with of cold ether (5 mL). Yield: 130 mg (70%). Mp: 149-151 °C. 1H NMR (CDCl3): p-cymene, δ 1.27 (d, J ) 7.0 Hz, 6H, CHMe2), 2.28 (s, 3H, Me), 2.87 (sept, J ) 7.0 Hz, 1H, CHMe2), 5.66 (d, J ) 5.8 Hz, 2H, MeC6H4CHMe2), 5.12 (d, J ) 5.8 Hz, 2H, MeC6H4CHMe2); P(OR)3, δ 3.80 (d, J ) 11.2 Hz, 9H, OMe). 13C{1H} NMR (CDCl ), p-cymene, δ 22.1 (s, C7), 18.1 (s, C1), 3 30.1 (s, C6), 101.8 (d, J ) 3.2 Hz, C2 or C5), 95.1 (d, J ) 2.2 Hz, C2 or C5), 81.2 (d, J ) 4.8 Hz, C3 or C4) 80.4 (d, J ) 5.7 Hz, C3 or C4); P(OR)3, δ 53.6 (d, J ) 5.7 Hz, P-OMe). Anal. Calcd for C13H23Cl2O3OsP: C, 30.1; H, 4.43. Found: C, 30.0; H, 4.31. Os(p-cymene)Cl2P(OEt)3 (7): Yellow powder, yield 140 mg (66%). Mp: 89-91 °C. 1H NMR (CDCl3): p-cymene, δ 1.27 (d, J ) 6.8 Hz, 6H, CHMe2), 2.27 (s, 3H, Me), 2.88 (sept, J ) 6.8 Hz, 1H, CHMe2), 5.63 (d, J ) 5.2 Hz, 2H, MeC6H4CHMe2), 5.48 (d, J ) 5.2 Hz, 2H, MeC6H4CHMe2); P(OR)3, δ 1.28 (t, J ) 6.8 Hz, 9H, CH3), 4.17 (m, J ) 6.8 Hz, 6H, OCH2). 13C{1H} NMR (CDCl3): p-cymene, δ 22.2 (s, C7), 18.0 (s, C1), 30.1 (s, C6), 101.1 and 94.5 (s, C2 and C5), 81.0 (d, J ) 5.3 Hz, C3 or C4), 80.3 (d, J ) 5.2 Hz, C3 or C4); P(OR)3, δ 16.1 (d, J ) 6.0 Hz, CH3), 62.4 (d, J ) 6.0 Hz, OCH2). Anal. Calcd for C16H29Cl2O3OsP: C, 34.2; H, 5.17. Found: C, 34.5; H, 5.19. Os(p-cymene)Cl2P(OnBu)3 (8): Orange powder, yield 145 mg (63%). Mp: 50-56 °C. 1H NMR (CDCl3): p-cymene, δ 1.25 (d, J ) 6.9 Hz, 6H, CHMe2), 2.24 (s, 3H, Me), 2.85 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.48 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2), 5.62 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2); P(OR)3, δ 0.94 (t, J ) 7.2 Hz, 9H, CH3), 4.08 (m, 6H, OCH2), 1.41 (m, 6H, CH2), 1.62 (m, 6H, CH2).13C{1H} NMR (CDCl3): p-cymene, δ 22.2 (s, C7), 17.9 (s, C1), 30.1 (s, C6), 100.5 and 93.8 (s, C2 and C5), 80.9 (d, J ) 5.6 Hz, C3 or C4) 80.6 (d, J ) 5.9 Hz, C3 or (45) Cabeza, J. A.; Maitlis, P. M. J. Chem. Soc., Dalton Trans. 1985, 573 (46) Werner, H.; Zenkert, K. J. Organomet. Chem. 1988, 345, 151. (47) Kiel, W. A.; Ball, R. G.; Graham, W. A. G. J. Organomet. Chem. 1990, 383, 481. (48) Arthur, T.; Stephenson, T. A. J Organomet. Chem. 1981, 208, 369.

3132

Organometallics, Vol. 15, No. 14, 1996

C4); P(OR)3, δ 13.7 (s, CH3), 66.1 (d, J ) 6.6 Hz, OCH2), 32.4 and 18.9 (s, CH2). Anal. Calcd for C22H41Cl2O3OsP: C, 40.9; H, 6.36. Found: C, 40.7; H, 6.14. Os(p-cymene)Cl2P(OiPr)3 (10): Orange powder, yield 190 mg (85%). Mp: 113-121 °C. 1H NMR (CDCl3): p-cymene, δ 1.26 (d, J ) 6.9 Hz, 6H, CHMe2), 2.23 (s, 3H, Me), 2.87 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.61 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2), 5.46 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2); P(OR)3, δ 1.30 (d, J ) 6.4 Hz, 18H, CH3), 4.94 (m, J ) 6.4 Hz, 3H, OCH). 13 C{1H} NMR (CDCl3): p-cymene, δ 22.3 (s, C7), 17.7 (s, C1), 30.2 (s, C6), 100.0 and 95.3 (s, C2 and C5), 80.6 (d, J ) 4.5 Hz, C3 or C4), 79.7 (d, J ) 6.3 Hz, C3 or C4); P(OR)3, δ 23.9 (d, J ) 3.8 Hz, CH3), 70.8 (d, J ) 7.0 Hz, OCH). Anal. Calcd for C19H35Cl2O3OsP: C, 37.8; H, 5.80. Found: C, 37.9; H, 5.53. Os(p-cymene)Cl2PMe2Ph (11): Orange powder, yield 180 mg (89%). Mp: 163-177 °C. 1H NMR (CDCl3): p-cymene, δ 1.12 (d, J ) 7.0 Hz, 6H, CHMe2), 1.96 (s, 3H, Me), 2.49 (sept, J ) 7.0 Hz, 1H, CHMe2), 5.34 (d, J ) 5.1 Hz, 2H, MeC6H4CHMe2), 5.24 (d, J ) 5.1 Hz, 2H, MeC6H4CHMe2); PR3, δ 1.85 (d, J ) 10.7 Hz, 6H, P-Me), 7.45 (m, 3H, Ph), 7.07 (t, J ) 8.5 Hz, 2H, Ph). 13C{1H} NMR (CDCl3), p-cymene, δ 21.8 (s, C7), 17.2 (s, C1), 29.7 (s, C6), 98.2 and 86.6 (s, C2 and C5), 80.2 (d, J ) 3.8 Hz, C3 or C4), 80.1 (d, J ) 4.5 Hz, C3 or C4); PR3, δ 11.4 (d, J ) 39.2 Hz, P-Me), 137.4 (d, J ) 49.0 Hz, ipso C6H5), 129.7, 129.1, and 128.1 (s, Ph). Anal. Calcd for C18H25Cl2OsP: C, 40.3; H, 4.69. Found: C, 40.6; H, 4.82. Os(p-cymene)Cl2PPh2Me (12): Orange-yellow powder, yield 200 mg (89%). Mp: 169-170 °C. 1H NMR (CDCl3): p-cymene, δ 0.92 (d, J ) 7.0 Hz, 6H, CHMe2), 1.98 (s, 3H, Me), 2.32 (sept, J ) 7.0 Hz, 1H, CHMe2), 5.45 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2), 5.39 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2); PR3, δ 1.96 (d, J ) 11.4 Hz, 3H, P-Me), 7.4-7.7 (m, 10H, Ph). 13C{1H} NMR (CDCl ): p-cymene, δ 21.7 (s, C7), 17.1 (s, C1), 3 29.6 (s, C6), 99.0 and 87.7 (s, C2 and C5), 80.7 and 77.9 (s, C3 and C4); PR3, δ 11.1 (d, J ) 39.6 Hz, P-Me), 134.7 (d, J ) 52.0 Hz, ipso C6H5), 132.7, 130.3 and 128.0 (s, Ph). Anal. Calcd for C23H27Cl2OsP: C, 46.4; H, 4.54. Found: C, 46.4; H, 4.46. Os(p-cymene)Cl2PPnBu3 (14): Orange powder, yield 196 mg (87%). Mp: 176-179 °C. 1H NMR (CDCl3): p-cymene, δ 1.29 (d, J ) 6.9 Hz, 6H, CHMe2), 2.22 (s, 3H, Me), 2.77 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.61 (d, J ) 5.6 Hz, 2H, MeC6H4CHMe2), 5.40 (d, J ) 5.6 Hz, 2H, MeC6H4CHMe2); PR3, δ 0.95 (t, J ) 6.9 Hz, 9H, CH3), 2.00 (m, 6H, P-CH2), 1.40 (m, 12H, CH2). 13C{1H} NMR (CDCl3): p-cymene, δ 22.3 (s, C7), 17.8 (s, C1), 30.3 (s, C6), 97.6 and 86.2 (s, C2 and C5), 80.1 (d, 3.1 Hz, C3 or C4), 76.6 (d, 6.5 Hz, C3 or C4); PR3, δ 13.6 (s, CH3), 23.7 (d, J ) 32.9 Hz, P-CH2), 25.1 (d, J ) 12.7 Hz, CH2), 24.2 (d, J ) 3.1, CH2). Anal. Calcd for C22H41Cl2OsP: C, 44.2; H, 6.87. Found: C, 44.3; H, 6.67. Os(p-cymene)Cl2PPiPr3 (17): Orange powder, yield 150 mg (71%). Mp: 142-146 °C. 1H NMR (CDCl3): p-cymene, δ 1.34 (d, J ) 7.5 Hz, 6H, CHMe2), 2.21 (s, 3H, Me), 2.83 (sept, J ) 7.5 Hz, 1H, CHMe2), 5.87 (d, J ) 5.5 Hz, 2H, MeC6H4CHMe2), 5.78 (d, J ) 5.5 Hz, 2H, MeC6H4CHMe2); PR3, δ 1.34 (dd, J(P,H) ) 12.9 Hz, J(H,H) ) 6.9 Hz, 18H, CH3), 2.83 (m, 3H, P-CH). 13C{1H} NMR (CDCl3): p-cymene, δ 23.0 (s, C7), 17.9 (s, C1), 30.7 (s, C6), 97.5 and 87.2 (s, C2 and C5), 79.9 (d, 3.0 Hz, C3 or C4), 76.7 (d, 3.0 Hz, C3 or C4); PR3, δ 20.1 (s, CH3), 25.3 (d, J ) 25.0 Hz, P-CH). Anal. Calcd for C19H35Cl2OsP: C, 41.1; H, 6.31. Found: C, 41.4; H, 6.17. Complexes 19-24 were similarly prepared by halogen exchange of the corresponding dichloride complex of III using NaI, for which a representative example is given for the PMe2Ph-substituted complex 19. Os(p-cymene)I2PMe2Ph (19). The dichloride 11 (50 mg, 0.09 mmol) and NaI (300 mg, 2.0 mmol) were stirred in MeOH (10 mL) overnight at room temperature. The solution was evaporated, redissolved in CH2Cl2 (10 mL), and filtered (Celite), and the volume of solvent was reduced to 2-3 mL. Pentane (15 mL) was added to precipitate a red powder, which was filtered out and washed with cold ether (5 mL). Yield:

Bell et al. 44 mg (65%). Mp: 158-170 °C. 1H NMR (CDCl3): p-cymene, δ 1.11 (d, J ) 6.9 Hz, 6H, CHMe2), 2.10 (s, 3H, Me), 2.88 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.27 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2), 5.21 (d, J ) 5.7 Hz, 2H, MeC6H4CHMe2); PR3, δ 2.24 (d, J ) 10.4 Hz, 6H, P-Me), 7.40-7.28 (m, 5H, Ph). 13C{1H} NMR (CDCl3): p-cymene, δ 22.3 (s, C7), 19.4 (s, C1), 31.2 (s, C6), 102.4 and 87.7 (s, C2 and C5), 81.5 (d, 2.9 Hz, C3 or C4), 77.1 (d, 4.6 Hz, C3 or C4); PR3, δ 18.7 (d, J ) 42.7 Hz, P-Me), 139.1 (d, J ) 45.5 Hz, ipso C6H5), 130.0 (s, Ph), 129.7 (d, J ) 7.9 Hz, Ph), 128.3 (d, J ) 9.3 Hz, Ph). Anal. Calcd for C18H25I2OsP: C, 30.2; H, 3.49. Found: C, 30.2; H, 3.58. MS: m/z 718 [M+] (192Os). Os(p-cymene)I2PnBu3 (20): Red powder, yield 42 mg (60%). Mp: 156-164 °C. 1H NMR (CDCl3): p-cymene, δ 1.28 (d, J ) 6.8 Hz, 6H, CHMe2), 2.38 (s, 3H, Me), 3.13 (sept, J ) 6.8 Hz, 1H, CHMe2), 5.12 (m, 4H, MeC6H4CHMe2); PR3, δ 0.95 (t, J ) 7.0 Hz, 9H, Me), 1.30-1.50 (m, 12H, CH2), 2.23 (m, 6H, P-CH2). 13C{1H} NMR (CDCl3): p-cymene, δ 22.5 (s, C7), 19.6 (s, C1), 31.5 (s, C6), 102.4 and 86.9 (s, C2 and C5), 80.9 and 76.2 (s, C3 or C4); PR3, δ 13.7 (s, CH3), 29.1 (d, J ) 33.6 Hz, P-CH2), 26.7 and 24.3 (s, CH2). Anal. Calcd for C22H41I2OsP: C, 33.9; H, 5.26. Found: C, 33.6; H, 5.20. MS: m/z 782 [M+] (192Os). Os(p-cymene)I2PPh2Me (21): Red powder, yield 45 mg (64%). Mp: 183-193 °C. 1H NMR (CDCl3): p-cymene, δ 0.91 (d, J ) 6.8 Hz, 6H, CHMe2), 2.18 (s, 3H, Me), 2.81 (sept, J ) 6.8 Hz, 1H, CHMe2), 5.50 (d, J ) 5.4 Hz, 2H, MeC6H4CHMe2), 5.37 (d, J ) 5.4 Hz, 2H, MeC6H4CHMe2); PR3, δ 2.57 (t, J ) 10.0 Hz, 3H, P-Me), 7.41-7.37 (m, 10H, Ph). 13C{1H} NMR (CDCl3): p-cymene, δ 21.9 (s, C7), 18.6 (s, C1), 31.0 (s, C6), 102.7 and 88.7 (s, C2 and C5), 81.7 (s, C3 or C4), 78.3 (d, 4.7 Hz, C3 or C4); PR3, δ 19.9 (d, J ) 44.4 Hz, P-Me), 136.8 (d, J ) 52.3 Hz, ipso C6H5), 133.3 (d, J ) 9.0 Hz, Ph), 130.2 (s, Ph), 127.9 (d, J ) 10.0 Hz, Ph). Anal. Calcd for C23H27I2OsP: C, 35.5; H, 3.47. Found: C, 35.8 H, 3.53. MS: m/z 780 [M+] (192Os). Os(p-cymene)I2PPh3 (22): Red powder, yield 47 mg (62%). Mp: 220-223 °C. 1H NMR (CDCl3): p-cymene, δ 1.25 (d, J ) 6.9 Hz, 6H, CHMe2), 2.01 (s, 3H, Me), 3.33 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.61 (d, J ) 5.8 Hz, 2H, MeC6H4CHMe2), 5.10 (d, J ) 5.6 Hz, 2H, MeC6H4CHMe2); PR3, δ 7.40-7.75 (m, 15H, Ph). 13C{1H} NMR (CDCl3): p-cymene, δ 22.8 (s, C7), 18.6 (s, C1), 31.7 (s, C6), 105.2 and 91.4 (s, C2 and C5), 81.9 and 80.3 (s, C3 and C4); PR3, δ 135.3 (d, J ) 53.8 Hz, ipso C6H5), 135.8 (d, J ) 9.0 Hz, Ph), 130.5 (s, Ph), 127.5 (d, J ) 10.1 Hz, Ph). Anal. Calcd for C28H29I2OsP: C, 40.0; H, 3.45. Found: C, 40.6; H, 3.52. MS: m/z 715 [M+ - I] (192Os). Os(p-cymene)I2PiPr3 (23): Red powder, yield 35 mg (53%). Mp: 151-163 °C. 1H NMR (CDCl3): p-cymene, δ 1.33 (d, J ) 7.3 Hz, 6H, CHMe2), 2.41 (s, 3H, Me), 3.17 (sept, J ) 7.3 Hz, 1H, CHMe2), 5.83 (m , 4H, MeC6H4CHMe2); PR3, δ 1.35 (dd, J(PH) ) 13.1 Hz, J(H,H) ) 7.4 Hz, 18H, CH3), 3.01 (m, 3H, P-CH). 13C{1H} NMR (CDCl3): p-cymene, δ 23.1 (s, C7), 19.4 (s, C1), 31.6 (s, C6), 102.2 and 87.2 (s, C2 and C5), 81.6 and 76.1 (s, C3 and C4); PR3, δ 21.1 (s, Me), 28.9 (d, J ) 26.0 Hz, P-CH). Anal. Calcd for C19H35I2OsP: C, 30.9; H, 4.74. Found: C, 30.9; H, 4.00. MS: m/z 613 [M+ - I] (192Os). Os(p-cymene)I2PCy3 (24): Red powder, yield 48 mg (62%). Mp: 143-151 °C. 1H NMR (CDCl3): p-cymene, δ 1.33 (d, J ) 6.9 Hz, 6H, CHMe2), 2.38 (s, 3H, Me), 3.19 (sept, J ) 6.9 Hz, 1H, CHMe2), 5.81 (m, 4H, J ) 5.9 Hz, MeC6H4CHMe2); PR3, δ 2.26 (m, 3H, P-CH), 2.20 (m, 6H, CH2), 1.81 (m, 12H, CH2), 1.40 (m, 12H, CH2). 13C{1H} NMR (CDCl3): p-cymene, δ 22.9 (s, C7), 19.3 (s, C1), 31.6 (s, C6), 102.4 and 87.0 (s, C2 and C5), 81.7 and 75.9 (s, C3 and C4); PR3, δ 38.9 (d, J ) 24 Hz, P-CH), 30.6 (s, CH2), 27.3 (d, J ) 9.6 Hz, CH2), 26.5 (s, CH2). No satisfactory microanalysis was obtained. Os(p-cymene)HClPMe3 (25). Complex 1 (90 mg, 0.19 mmol) and Et3N (0.5 mL, 0.5 mmol) were stirred at 80 °C in ethanol (20 mL) for 5 h. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral

(η6-Arene)osmium(II) Complexes grade (IV) alumina and diethyl ether as eluant, giving a yellow powder. Yield: 50 mg (60%). Mp: 121-123 °C. 1H NMR (C6D6): p-cymene, δ 1.08 (d, J ) 6.7 Hz, 3H, CHMe2), 1.11 (d, J ) 6.7 Hz, 3H, CHMe2), 2.02 (s, 3H, Me), 2.13 (sept, J ) 6.7 Hz, 1H, CHMe2), 5.23 (d, 1H, J ) 5.2 Hz, MeC6H4CHMe2), 4.98 (d, 1H, J ) 5.2 Hz, MeC6H4CHMe2), 4.46 (d, 1H, J ) 5.6 Hz, MeC6H4CHMe2), 4.25 (d, 1H, J ) 5.6 Hz, MeC6H4CHMe2); PR3, δ 1.34 (d, J ) 10.0 Hz, 9H, P-Me); -8.63 (d, J ) 45.4 Hz, 1H, Os-H). 13C{1H} NMR (C6D6): p-cymene, δ 25.1 and 23.2 (s, C7), 19.6 (s, C1), 32.4 (s, C6), 96.4 and 93.5 (s, C2 and C5), 78.8, 78.5, 78.6, 72.0 (s, C3 and C4); PR3, δ 20.8 (d, J ) 37.6 Hz, P-Me). MS: m/z 437 [M+ - H] (192Os, 35Cl), C13H24ClOsP. ν(Os-H) (CH2Cl2): 2056 cm-1. Os(p-cymene)HClPMe2Ph (26). Complex 11 (90 mg, 0.17 mmol) and Et3N (0.5 mL, 5.0 mmol) were stirred at 80 °C in ethanol (20 mL) for 5 h. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral grade (IV) alumina and diethyl ether as eluant, giving a yellow oil. Yield: 52 mg (62%). 1H NMR (C6D6): p-cymene, δ 1.02 (d, J ) 6.8 Hz, 3H, CHMe2), 1.08 (d, J ) 6.8 Hz, 3H, CHMe2), 1.87 (s, 3H, Me), 2.00 (sept, J ) 6.8 Hz, 1H, CHMe2), 5.17 (d, 1H, J ) 5.6 Hz, MeC6H4CHMe2); 4.81 (d, 1H, J ) 5.3 Hz, MeC6H4CHMe2), 4.42 (d, 1H, J ) 5.6 Hz, MeC6H4CHMe2), 4.19 (d, 1H, J ) 5.3 Hz, MeC6H4CHMe2); PR3, δ 1.84 (d, J ) 10.3 Hz, 3H, P-Me), 1.64 (d, J ) 10.0 Hz, 3H, P-Me); 7.05-7.50 (m, 5H, Ph); -8.59 (d, J ) 45.2 Hz, 1H, Os-H). 13C{1H} NMR (C6D6): p-cymene, δ 24.9 and 23.1 (s, C7), 19.2 (s, C1), 31.9 (s, C6), 98.3 and 93.5 (s, C2 and C5), 80.2, 79.8, 79.0, 72.5 (s, C3 and C4); PR3, δ 21.2 (d, J ) 41.8 Hz, P-Me), 17.6 (d, J ) 37.1 Hz, P-Me), 141.4 (d, J ) 46.9 Hz, ipso C6H5), 130.4 and 129.6 (s, Ph), remainder under solvent resonance. MS: m/z 500 [M+] (192Os, 35Cl), C18H26ClOsP. ν(Os-H) (CH2Cl2): 2022 cm-1. Os(p-cymene)HClPnBu3 (27). Complex 14 (90 mg, 0.15 mmol) and Et3N (0.5 mL, 5.0 mmol) were stirred at 80 °C in ethanol (20 mL) for 5 h. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral grade (IV) alumina and diethyl ether as eluant, giving a yellow oil. Yield: 55 mg (65%). 1H NMR (C6D6): p-cymene, δ 1.21 (d, J ) 7.3 Hz, 3H, CHMe2), 1.13 (d, J ) 6.7 Hz, 3H, CHMe2), 2.06 (s, 3H, Me), 2.14 (m, 1H, CHMe2), 5.45 (d, 1H, J ) 5.7 Hz, MeC6H4CHMe2); 5.22 (d , 1H, J ) 5.1 Hz, MeC6H4CHMe2), 4.52 (d, 1H, J ) 5.7 Hz, MeC6H4CHMe2), 4.17 (d, 1H, J ) 5.1 Hz, MeC6H4CHMe2); PR3, δ 0.9 (t, J ) 7.1 Hz, 9H, CH3), 1.73 (m, 6H, P-CH2); 1.32 (m, 12H, CH2); -8.61 (d, J ) 43.2 Hz, 1H, Os-H). 13C{1H} NMR (C6D6): p-cymene, δ 24.9 and 22.0 (s, C7), 19.1 (s, C1), 32.5 (s, C6), 94.8 and 93.0 (s, C2 and C5), 81.1 , 81.2, 78.6, 69.7 (s, C3 and C4); PR3, δ 14.5 (s, CH3), 28.4 (d, J ) 33.2 Hz, P-CH2), 26.7 and 25.1 (s, CH2). MS: m/z 564 [M+] (192Os, 35Cl), C22H42ClOsP. ν(Os-H) (CH2Cl2): 2060 cm-1. Os(p-cymene)HClPPh2Me (28). Complex 12 (90 mg, 0.15 mmol) and Et3N (0.5 mL, 5.0 mmol) were stirred at 80 °C in ethanol (20 mL) for 5 h. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral grade (IV) alumina and diethyl ether as eluant, giving a yellow oil. Yield: 55 mg (65%). 1H NMR (C6D6): p-cymene, δ 0.96 (d, J ) 6.8 Hz, 3H, CHMe2), 1.05 (d, J ) 7.2 Hz, 3H, CHMe2), 1.87 (s, 3H, Me), 1.81 (m, 1H, CHMe2), 5.27 (d, 1H, J ) 5.6 Hz, MeC6H4CHMe2); 4.96 (d , 1H, J ) 5.2 Hz, MeC6H4CHMe2), 4.49 (d, 1H, J ) 5.6 Hz, MeC6H4CHMe2), 4.09 (d, 1H, J ) 5.2 Hz, MeC6H4CHMe2); PR3, δ 2.18 (d, J ) 9.8 Hz, 3H, P-Me) 6.90-7.71 (m, 10H, Ph); -8.40 (d, J ) 44.6 Hz, 1H, Os-H). 13C{1H} NMR (C D ): p-cymene, δ 24.9 and 22.9 (s, C7), 18.7 6 6 (s, C1), 31.5 (s, C6), 98.4 and 94.5 (s, C2 and C5), 81.4 (s, C3 or C4), 81.1 (d, J ) 5.3 Hz, C3 or C4), 79.2 (s, J ) 6.0 Hz, C3 or C4), 72.5 (s, C3 or C4); PR3, δ 20.3 (d, J ) 41.5 Hz, P-Me), 140.7 (d, J ) 53.6 Hz, ipso C6H5), 137.9 (d, J ) 47.2 Hz, ipso C6H5), 133.1, 132.5, 130.1 and 129.6 (s, C6H5), remainder under

Organometallics, Vol. 15, No. 14, 1996 3133 solvent resonance. MS: m/z 562 [M+] (192Os, 35Cl), C23H28ClOsP. ν(Os-H) (CH2Cl2): 2059 cm-1. Os(p-cymeneHClPiPr3 (30). Complex 17 (100 mg, 0.18 mmol) and Et3N (0.5 mL, 0.5 mmol) were stirred at 80 °C in ethanol (20 mL) for 10 min. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral grade (IV) alumina and diethyl ether as eluant, giving a yellow oil. Yield: 84 mg (90%). 1H NMR (C6D6): p-cymene, δ 1.99 (s, 3H, Me), 2.13 (sept, J ) 7.0 Hz, 1H, CHMe2), 5.61 (d, 1H, J ) 5.8 Hz, MeC6H4CHMe2); 5.50 (d, 1H, J ) 5.3 Hz, MeC6H4CHMe2), 4.41 (d, 1H, J ) 5.8 Hz, MeC6H4CHMe2), 4.19 (d, 1H, J ) 5.8 Hz, MeC6H4CHMe2), CHMe2 under PiPr3 resonance; PR3, δ 1.04-1.23 (m, 24H, CHMe2 and P-CHMe2), 2.25 (m, 3H, P-CH); -8.61 (d, J ) 41.5 Hz, 1H, Os-H). 13C{1H} NMR (C6D6): p-cymene, δ 21.6 and 22.1 (s, C7), 18.9 (s, C1), 30.2 (s, C6), 95.8 and 91.0 (s, C2 and C5), 82.5, 81.0, 78.3 and 70.2 (s, C3 and C4); PR3, δ 19.9 (d, J ) 12.4 Hz, Me), 27.0 (d, J ) 51.9 Hz, P-CH). MS: m/z 522 [ M+] (192Os, 35Cl), C19H36ClOsP. ν(Os-H) (CH2Cl2): 2084 cm-1. Os(p-cymene)HClPCy3 (31). Complex 18 (100 mg, 0.15 mmol) and Et3N (0.5 mL, 5.0 mmol) were stirred at room temperature in ethanol (20 mL) for 10 min. The cooled solution was filtered (Celite) and the volume of solvent reduced to 2-3 mL. The product was purified by column chromatography using neutral grade (IV) alumina and diethyl ether as eluant, giving a yellow powder. Yield: 50 mg (53%). Mp: 153-159 °C. 1H NMR (C6D6): p-cymene, δ 1.16 (d, J ) 6.9 Hz, 3H, CHMe2), 1.25 (d, J ) 6.9 Hz, 3H, CHMe2), 2.02 (s, 3H, Me), CHMe2 under PCy3 resonance, 5.68 (d, 1H, J ) 5.4 Hz, MeC6H4CHMe2); 5.61 (d, 1H, J ) 5.0 Hz, MeC6H4CHMe2), 4.40 (d, 1H, J ) 5.4 Hz, MeC6H4CHMe2), 4.26 (d, 1H, J ) 5.0 Hz, MeC6H4CHMe2); PR3, δ 2.21 (m, 3H, P-CH), 2.14 (m, 6H, CH2); 1.95 (m, 4H, CH2); 1.76 (m, 6H, CH2); 1.54 (m, 12H, CH2); -8.45 (d, J ) 41.8 Hz, 1H, Os-H). 13C{1H} NMR (C6D6): p-cymene, δ 25.6 and 21.4 (s, C7), 18.9 (s, C1), 31.7 (s, C6), 96.0 and 89.4 (s, C2 and C5), 82.7, 79.1, 78.3, 70.1 (s, C3 and C4); PR3, δ 37.0 (d, J ) 26.8 Hz, P-CH), 30.1, 28.2, and 27.4 (CH2). MS: m/z 641 [M+ - H] (192Os, 35Cl), C28H48ClOsP. ν(Os-H) (CH2Cl2): 2088 cm-1. Os(C6H6)Cl2PMe3 (32). [Os(C6H6)Cl2]2 (300 mg, 0.44 mmol) was stirred at 60 °C with PMe3 (0.3 mL, 4 mmol) in toluene for 2 h. Additional PMe3 (0.3 mL) was added and stirring continued for a further 2 h. The cooled solution was filtered (Celite) and washed with CD2Cl2 until the filtrate was clear. The solution was evaporated and redissolved in CH2Cl2 (5 mL), and pentane (20 mL) was added to obtain an orange precipitate, which was filtered off and washed with cold ether (5 mL). Yield: 210 mg (57%). Mp: decomposes at 220-230 °C. 1H NMR (CDCl3): δ 1.67 (d, 9H, J ) 11.1 Hz, P-Me), 5.73 (s, 6H, C6H6). 13C{1H} NMR (CDCl3): δ 15.0 (d, J ) 38.3 Hz, P-Me), 78.6 (d, J ) 2.8 Hz, C6H6). Anal. Calcd for C9H15Cl2OsP: C, 26.0; H, 3.61. Found: C, 26.0; H, 3.95. MS: m/z 416 [M+] (192Os, 35Cl). [Os(C6H5Me)Cl2]2. 2-Methylcyclohexa-1,4-diene 49 (1.0 g, 100 mmol) and OsCl3 (500 mg, 17 mmol) were stirred in ethanol (15 mL) for 3 days at 80 °C. The cooled solution was filtered (Celite) and washed with hot CDCl3, and the volume was reduced to 2-3 mL. Pentane (20 mL) was added to precipitate a yellow powder, which was filtered off and washed with cold ether (5 mL). Yield: 520 mg (87%). Mp: slowly decomposes above 170 °C. 1H NMR (CDCl3): δ 2.24 (s, 6H, Me), 6.05 (d, J ) 5.0 Hz, 4H, CH), 6.21 (t, 2H, J ) 4.8 Hz, CH), 6.28 (m, 4H, CH). 13C DEPT NMR (CDCl3): δ 19.7 (s, Me), 78.8, 72.2 and 70.8 (s, CH) (quaternary C not visible). Anal. Calcd for C14H16Cl4Os2: C, 23.8; H, 2.27. Found: C, 24.8; H, 2.93. MS: m/z 354 [M+/2] (192Os, 35Cl). (49) Krapcho, A. P.; Bothner-By, A. A. J. Am. Chem. Soc. 1959, 81, 3658. (50) Johnson, C. K. ORTEPII; Report ORNL-5138; Oak Ridge National Laboratory: Oak Ridge, TN, 1976.

3134

Organometallics, Vol. 15, No. 14, 1996

Os(C6H5Me)Cl2PMe3 (33). [Os(C6H5Me)Cl2]2 (150 mg, 0.2 mmol) and P(Me)3 (0.2 mL, 2.6 mmol) were stirred in CH2Cl2 (15 mL) for 1 h at 40 °C. The cooled solution was filtered (Celite) and the volume reduced to 2-3 mL. Pentane (15 mL) was added to precipitate a yellow powder, which was filtered off and washed with of cold ether (5 mL). Yield: 133 mg (73%). Mp: 209-215 °C. 1H NMR (CDCl3): δ 2.28 (s, Me), 5.67 (m, 2H, CH), 5.47 (m, 2H, CH), 5.42 (t, 1H, CH), 1.68 (d, J ) 10.5 Hz, 9H, PMe3). 13C{1H} NMR (CDCl3): δ 18.5 (s, 3H, Me), 15.4 (d, J ) 38.2 Hz, PMe3), 99.9 (s, quaternary), 78.7, 77.9, and 69.6 (s, CH). Anal. Calcd for C10H17Cl2OsP: C, 28.0; H, 3.96. Found: C, 28.3; H, 3.57. [Os(C6H5Et)Cl2]2. 2-Ethylcyclohexa-1,4-diene49 (1.0 g, 100 mmol) and OsCl3 (500 mg, 17 mmol) were stirred in ethanol (15 mL) for 3 days at 80 °C. The cooled solvent was filtered (Celite) and washed with hot CDCl3, and the volume of solvent was reduced to 2-3 mL. Pentane (20 mL) was added to precipitate a yellow powder which was filtered off and washed with cold ether (5 mL). Yield: 560 mg (90%). Mp: decomposes 200-225 °C. 1H NMR (CDCl3): δ 1.29 (t, J ) 7.4 Hz, 6H, CH3), 2.55 (q, J ) 7.4 Hz, 4H, CH2), 6.08 (d, J ) 4.4 Hz, 4H, CH), 6.20 (t, J ) 3.9 Hz), 2H, CH), 6.28 (m, 4H, CH). 13C{1H} NMR (CDCl3): δ 14.0 (s, CH3), 27.0 (s, CH2), 94.2 (s, quaternary), 76.5, 71.7, and 71.4 (s, CH). Anal. Calcd for C16H20Cl4Os2: C, 26.2; H, 2.72. Found: C, 26.4; H, 3.13. MS: m/z 368 [M+/ 2] (192Os, 35Cl). Os(C6H5Et)Cl2PMe3 (34). [Os(C6H5Et)Cl2]2 (200 mg, 0.27 mmol) and P(Me)3 (0.2 mL, 2.6 mmol) were stirred in CH2Cl2 (15 mL) for 1 h at 40 °C. The cooled solution was filtered (Celite) and the volume reduced to 2-3 mL. Pentane (15 mL) was added to precipitate a yellow powder, which was filtered out and washed with cold ether (5 mL). Yield: 205 mg (84%). Mp: 176-179 °C. 1H NMR (CDCl3): δ 1.31 (t, J ) 7.4 Hz, 3H, CH3), 2.57 (q, J ) 7.4 Hz, 2H, CH2), 5.67 (m, 2H, CH), 5.52 (m, 2H, CH), 5.31 (t, J ) 5.3 Hz, 1H, CH), 1.66 (d, J ) 10.7 Hz, 9H, PMe3). 13C{1H} NMR (CDCl3): δ 13.9 (s, CH3), 25.7 (s, CH2), 15.4 (d, J ) 38.4 Hz, PMe3), 103.6 (s, quaternary), 78.5, 77.5, and 70.3 (s, CH). Anal. Calcd for C11H19Cl2OsP: C, 29.8; H, 4.29. Found: C, 29.8; H, 3.99. [Os(C6H5iPr)Cl2]2. 2-Isopropylcyclohexa-1,4-diene49 (1.2 g, 100 mmol) and OsCl3 (500 mg, 17 mmol) were stirred in ethanol (15 mL) for 3 days at 80 °C. The cooled solvent was filtered (Celite) and washed with hot CDCl3, and the volume of solvent was reduced to 2-3 mL. Pentane (20 mL) was added to precipitate a yellow powder which was filtered off and washed with cold ether (5 mL). Yield: 460 mg (72%). Mp: slowly decomposes above 170 °C. 1H NMR (CDCl3): δ 1.31 (s, J ) 6.2, 12H, CHMe2), 2.82 (sept, J ) 6.2, 2H, CHMe2), 6.19 (m, 6H, CH), 5.46 (m, 4H, CH). 13C{1H} NMR (CDCl3): δ 22.2 (s, CH3), 31.5 (s, CH), 69.3 (s, quaternary), 75.0, 73.0, and 71.8 (s, CH). Anal. Calcd for C18H24Cl4Os2: C, 28.4; H, 3.15. Found: C, 28.8; H, 2.85. MS: m/z 382 [M+/2] (192Os, 35Cl). Os(C6H5iPr)Cl2PMe3 (35). [Os(C6H5iPr)Cl2]2 (200 mg, 0.26 mmol) and PMe3 (0.2 mL, 2.6 mmol) were stirred in CH2Cl2 (15 mL) for 1 h at 40 °C. The cooled solution was filtered (Celite) and the volume reduced to 2-3 mL. Pentane (15 mL) was added to precipitate a yellow powder, which was filtered out and washed with cold ether (5 mL). Yield: 220 mg (92%). Mp: decomposes 180-195 °C. 1H NMR (CDCl3): δ 1.32 (s, J ) 6.8, 6H, CHMe2), 2.90 (sept, J ) 6.8, 1H CHMe2), 5.59 (m, 4H, CH), 5.46 (t, J ) 5.3 Hz, 1H, CH), 1.66 (d, J ) 10.6 Hz, 9H, PMe3). 13C{1H} NMR (CDCl3): δ 22.0 (s, CH3), 30.6 (s, CH), 15.2 (d, J ) 38.0 Hz, PMe3), 107.7 (s, quaternary), 77.3, 77.0, and 70.3 (s, CH). Anal. Calcd for C12H21Cl2OsP: C, 31.5; H, 4.60. Found: C, 31.4; H, 4.31. [Os(C6H5tBu)Cl2]2. 2-tert-Butylcyclohexa-1,4-diene49 (1.3 g, 100 mmol) and OsCl3 (500 mg, 17 mmol) were stirred in ethanol (15 mL) for 3 days at 80 °C. The cooled solvent was filtered (Celite) and washed with hot CDCl3, and the volume was reduced to 2-3 mL. Pentane (20 mL) was added to precipitate a yellow powder, which was filtered off and washed

Bell et al. with cold ether (5 mL). Yield: 600 mg (90%). Mp: slowly decomposes above 180 °C. 1H NMR (CDCl3): δ 1.40 (s, 18H, CMe3), 6.57 (d, J ) 5.3 Hz, 4H, CH), 5.57 (m, 6H, CH). 13C{1H} NMR (CDCl3): δ 30.6 (s, CMe3), 35.3 (s, CMe3), 95.3 (s, quaternary), 77.6, 74.4, and 70.9 (s, CH). Anal. Calcd for C20H28Cl4Os2: C, 30.4; H, 3.54. Found: C, 30.5; H, 4.09. MS: m/z 396 [M+/2] (192Os, 35Cl). Os(C6H5tBu)Cl2PMe3 (36). [Os(C6H5tBu)Cl2]2 (200 mg, 0.27 mmol) and P(Me)3 (0.2 mL, 2.6 mmol) were stirred in CH2Cl2 (15 mL) for 1 h at 40 °C. The cooled solution was filtered (Celite) and the volume reduced to 2-3 mL. Pentane (15 mL) was added to precipitate a yellow powder, which was filtered out and washed with cold ether (5 mL). Yield: 210 mg (83%). Mp: decomposes at 180-200 °C. 1H NMR (CDCl3): δ 1.45 (s, 9H, CMe3), 5.95 (d, J ) 5.8 Hz, 2H, CH), 5.65 (m, 2H, CH), 5.46 (t, J ) 5.2 Hz, 1H, CH), 1.63 (d, J ) 10.6 Hz, 9H, P-Me). 13C{1H} NMR (CDCl3): δ 30.8 (s, CMe3), 35.2 (s, CMe3), 108.4 (d, J ) 38.4 Hz, quaternary), 80.6 (d, J ) 4.5 Hz, CH), 71.9 and 71.8 (s, CH). Anal. Calcd for C12H23Cl2OsP: C, 33.1; H, 4.88. Found: C, 33.0; H, 4.68. Os(C6Me6)Cl2PMe3 (37). [Os(C6Me6)Cl2]2 (200 mg, 0.24 mmol) and P(Me)3 (0.2 mL, 2.6 mmol) were stirred in CH2Cl2 (15 mL) for 1 h at 40 °C. The cooled solution was filtered (Celite) and the volume reduced to 2-3 mL. Pentane (15 mL) was added to precipitate a yellow powder, which was filtered out and washed with cold ether (5 mL). Yield: 132 mg (62%). Mp: decomposes at 209-219 °C. 1H NMR (CDCl3): δ 2.08 (s, 18H, Me), 1.53 (d, J ) 10.3 Hz, 9H, P-Me). 13C{1H} NMR (CDCl3): δ 16.0 (s, CMe), 87.3 (d, J ) 3.7 Hz, CMe), 14.7 (d, J ) 37.0, P-Me). Anal. Calcd for C15H27Cl2OsP: C, 36.1; H, 5.41. Found: C, 35.6; H, 4.82. MS: m/z 500 [M+] (192Os, 35Cl). X-ray Structure Determinations. The structures of 1, 6, 13, and 15 were determined at 173 K. All measurements were made on a Rigaku AFC5R diffractometer using graphitemonochromated Mo KR radiation (λ ) 0.710 69 Å) and a 12 kW rotating anode generator. For each compound, the intensities of three standard reflections were measured after every 150 reflections and remained stable throughout the data collection. The intensities were corrected for Lorentz and polarization effects. Semi-empirical absorption corrections, based on ψ-scans of several reflections, were applied to 1 and 13.51 An analytical absorption correction was used for 6,52 and an empirical absorption correction using the program DIFABS53 was used for 15 because other absorption correction methods produced unsatisfactory results. Equivalent reflections were merged. Data collection and refinement parameters are given in Table 6. The structures were solved by Patterson methods using SHELXS8654 and DIRDIF92,55 which revealed the positions of the heavy atoms. All remaining non-hydrogen atoms were located in Fourier expansions of the Patterson solutions. The non-hydrogen atoms were refined anisotropically. The H atoms were fixed in geometrically calculated positions [d(CH) ) 0.95 Å], and they were assigned fixed isotropic temperature factors with a value of 1.2Ueq of the parent C atom. The orientations of the methyl group H atoms were based on peaks located in difference electron density maps. Refinement of each structure was carried out on F using full-matrix leastsquares procedures, which minimized the function ∑w(|Fo| |Fc|)2. The weighting scheme was based on counting statistics and included a factor to downweight the intense reflections. Plots of ∑w(|Fo| - |Fc|)2 versus |Fo|, reflection order in data (51) North, A. C. T.; Phillips, D. C.; Mathews, F. S. Acta Crystallogr. 1968, A24, 351. (52) De Meulenaer, J.; Tompa, H. Acta Crystallogr. 1965, 19, 1014. (53) Walker, N.; Stuart, D. Acta Crystallogr. 1983, A39, 158. (54) Sheldrick, G. M. SHELXS-86. Acta Crystallogr. 1990, A46, 467. (55) Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.; Garcı´a-Granda, S.; Smits, J. M. M.; Smykalla, C. DIRDIF-92. The DIRDIF program system; Technical Report of the Crystallography Laboratory; University of Nijmegen: Nijmegen, The Netherlands, 1992.

(η6-Arene)osmium(II) Complexes

Organometallics, Vol. 15, No. 14, 1996 3135 Table 6. Crystallographic Data for 1, 6, 13, and 15

crystallized from empirical formula fw cryst color, habit cryst dimens, mm temp, K cryst system space group Z reflcns for cell determ 2θ range for cell determ, deg a, Å b, Å c,Å R, deg β, deg γ, deg V, Å3 F(000) Dx, g cm-3 µ(Mo KR), mm-1 scan type 2θ(max), deg min and max transm factors tot. reflcns measd sym indepdt reflcns Rmerge reflcns used [I > 2σ(I)] params refined R Rw weights: p in w ) [σ2(Fo) + (pFo)2]-1 goodness of fit final ∆max/σ max and min ∆F, e Å-3

1

6

13

15

methanol C13H23Cl2OsP 471.40 orange, thin prism 0.08 × 0.25 × 0.43 173(1) monoclinic P21/n 8 25 39-40 10.269(6) 27.168(7) 13.002(4) 90 112.37(3) 90 3354(2) 1808 1.867 7.991 ω 55 0.449, 1.000 8276 (+h,+k,(l) 7679 0.020 6097 307 0.0296 0.0260 0.005 1.364 0.0007 1.48, -0.89

hexane/ethyl acetate C13H23Cl2O3OsP 519.40 orange, needle 0.10 × 0.21 × 0.39 173(1) monoclinic P21/c 4 25 39-40 13.329(2) 7.286(2) 17.480(2) 90 91.48(1) 90 1697.1(5) 1000 2.033 7.921 ω /2θ 60 0.096, 0.378 5515 (+h,+k,(l) 4949 0.054 3665 181 0.0335 0.0288 0.005 1.330 0.0008 1.26, -1.03

ethanol C31H35Cl2OsP 699.70 orange, prism 0.15 × 0.25 × 0.30 173(1) triclinic P1 h 2 25 39-40 10.381(1) 14.610(1) 9.277(1) 96.860(9) 97.01(1) 88.631(8) 1386.4(3) 692 1.676 4.865 ω /2θ 60 0.479, 1.000 8494 (+h,(k,(l) 8088 0.017 7341 316 0.0233 0.0228 0.005 1.367 0.002 1.72, -1.13

CH2Cl2 C28H29Cl2OsP.CH2Cl2 742.55 orange, irregular prism 0.38 × 0.38 × 0.43 173(1) triclinic P1h 2 25 39-40 12.625(2) 13.509(2) 9.361(2) 109.38(1) 102.27(2) 99.67(1) 1421.4(5) 728 1.735 4.932 ω /2θ 60 0.818, 1.098 8592 (+h,(k,(l)) 8249 0.021 7327 316 0.0285 0.0325 0.01 1.549 0.001 1.29, -1.25

collection, (sin θ)/λ, and various classes of indices showed no unusual trends. Corrections for secondary extinction were not applied. Neutral atom scattering factors for non-hydrogen atoms were taken from Maslen, Fox, and O’Keefe,56a and the scattering factors for H atoms were taken from Stewart, Davidson, and Simpson.57 Anomalous dispersion effects were included in Fc;58 the values for f ′ and f ′′ were those of Creagh and McAuley.56b All calculations were performed using the TEXSAN crystallographic software package.59 The structure of 1 has two molecules in the asymmetric unit; however, there are no major differences in their conformation other than a small rotation of the PMe3 group. Some of the terminal methyl groups, especially in molecule A, are undergoing strong thermal motion (rotation about the C(1)-C(7)

bond) or are slightly disordered. However, no attempt was made to resolve any potential disorder. As a result, the C(7)C(8) and C(7)-C(9) bond lengths appear to be shorter than they probably are. The crystals of 15 contain molecules of CH2Cl2 in a ratio of 1:1 with the osmium complex. The residual electron density of between 1.2 and 1.7 e Å-3 in each structure was located near the Os atoms and near the Cl atoms of the solvent molecule in 15. The latter peaks might be due to slight disorder of the solvent molecule.

(56) International Tables for Crystallography; Wilson, A. J. C., Ed.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1992; Vol. C: (a) Maslen, E. N.; Fox, A. G.; O’Keefe, M. A. Table 6.1.1.1, pp 477486; (b) Creagh, D. C.; McAuley, W. J. Table 4.2.6.8, pp 219-222. (57) Stewart, R. F.; Davidson, E. R.; Simpson, W. T. J. Chem. Phys. 1965, 42, 3175. (58) Ibers, J. A.; Hamilton, W. C. Acta Crystallogr. 1964, 17, 781. (59) TEXSAN. Single Crystal Structure Analysis Software, Version 5.0; Molecular Structure Corp.: The Woodlands, TX, 1989.

Supporting Information Available: Listings of fractional atomic coordinates and equivalent isotropic temperature factors, interatomic distances, bond angles, torsion angles, anisotropic temperature factors, and hydrogen atom coordinates for compounds 1, 6, 13, and 15 (24 pages). Ordering information is given on any current masthead page.

Acknowledgment. We thank the Swiss National Science Foundation and Dr. Helmut Legerlotz-Stiftung for financial support and Dr. A. Hafner (Ciba-Geigy, Marly, Switzerland) for providing stimulating discussions.

OM960053I